I'm backing up a volume on a Gluster four node pool (distributed, replicated with replica count 2) using Bareos 16.2.4 on it's own server. The backup is put on another Gluster volume.
The initial full backup worked (took 22 hours). I then added a test file and the incremental worked. A day or so later I copied files from a windows machine to the Gluster volume. The Gluster volume is also shared via Samba 4. The files and directories were added to the volume. I then ran an incremental and only the test file was found. I deleted the test file and when I run an incremental or differential backup the test file is still listed as new. I delete the test file and get the same results. If I run the bareos glusterfind wrapper specifying a differential backup and since the last full backup the only file found is the test file which no longer exists. None of the additional files show up but they are in the volume and can be listed. If I go to the Gluster server that Bareos runs the script on and run the glusterfind directly with this command I still get only the test file listed. glusterfind query --since-time=1495843200 dev-projects dev-projects.out If I run find on the volume to get all files since that date and time all the new files are listed. It appears that somewhere, somehow gluster doesn't really walk the tree but is pulling the information out of it's hat from somewhere. I've stopped and started the volume, restarted gluster on all nodes and the query still returns the non-existent file as the ONLY file. I'm at a loss to explain this. I've gone through the glusterfind document on the gluster site but didn't find anything that might explain this behavior. Any suggestions are welcome. -- You received this message because you are subscribed to the Google Groups "bareos-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. For more options, visit https://groups.google.com/d/optout.
